{
 "runs": [
  {
   "month": "2026-08",
   "label": "August 2026",
   "ts": "2026-08-10T08:22:31+00:00",
   "scores": {
    "nvidia/nemotron-3-ultra-550b-a55b:free": {
     "rank": 1,
     "composite": 9.9,
     "A": 9.8,
     "B": 10.0,
     "C": 10.0,
     "tokens": 2086,
     "tok_s": 17.1,
     "cost": 0.0,
     "dnf": []
    },
    "google/gemma-4-26b-a4b-it:free": {
     "rank": 2,
     "composite": 9.8,
     "A": 10.0,
     "B": 9.4,
     "C": 10.0,
     "tokens": 1392,
     "tok_s": 23.4,
     "cost": 0.0,
     "dnf": []
    },
    "poolside/laguna-s-2.1:free": {
     "rank": 3,
     "composite": 9.1,
     "A": 9.8,
     "B": 9.4,
     "C": 8.0,
     "tokens": 908,
     "tok_s": 28.2,
     "cost": 0.0,
     "dnf": []
    },
    "poolside/laguna-xs-2.1:free": {
     "rank": 4,
     "composite": 8.8,
     "A": 9.0,
     "B": 9.4,
     "C": 8.0,
     "tokens": 2957,
     "tok_s": 41.2,
     "cost": 0.0,
     "dnf": []
    },
    "inclusionai/ling-3.0-tiny:free": {
     "rank": 5,
     "composite": 3.8,
     "A": 9.0,
     "B": 0.0,
     "C": 7.5,
     "tokens": 1614,
     "tok_s": 92.0,
     "cost": 0.0,
     "dnf": [
      "B"
     ]
    },
    "google/gemma-4-31b-it:free": {
     "rank": 6,
     "composite": 0.0,
     "A": 0.0,
     "B": 0.0,
     "C": 0.0,
     "tokens": 0,
     "tok_s": 0,
     "cost": 0.0,
     "dnf": [
      "A",
      "B",
      "C"
     ]
    }
   },
   "external": {
    "nvidia/nemotron-3-ultra-550b-a55b:free": {
     "name": "NVIDIA: Nemotron 3 Ultra (free)",
     "context": 1000000,
     "created": 1780551208,
     "description": "NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...",
     "hf": {
      "repo": "nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16",
      "downloads": 494365,
      "likes": 312
     },
     "arena": {
      "elo": 1426.0,
      "votes": 10564,
      "set": "text"
     },
     "aa_index": 38
    },
    "google/gemma-4-26b-a4b-it:free": {
     "name": "Google: Gemma 4 26B A4B  (free)",
     "context": 262144,
     "created": 1775227989,
     "description": "Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference — delivering near-31B quality at...",
     "hf": {
      "repo": "google/gemma-4-26B-A4B-it",
      "downloads": 10188506,
      "likes": 1366
     },
     "arena": {
      "elo": 1438.2,
      "votes": 5755,
      "set": "text"
     }
    },
    "poolside/laguna-s-2.1:free": {
     "name": "Poolside: Laguna S 2.1 (free)",
     "context": 262144,
     "created": 1784652683,
     "description": "Laguna S 2.1 is the latest coding agent model from [Poolside](<https://poolside.ai/>). Laguna S 2.1 is a 118B total parameter model with 8B active parameters, scoring 70.2% on Terminal-Bench 2.1 and...",
     "hf": null
    },
    "poolside/laguna-xs-2.1:free": {
     "name": "Poolside: Laguna XS 2.1 (free)",
     "context": 262144,
     "created": 1783002429,
     "description": "Laguna XS 2.1 is the latest coding agent model in the 33B-A3B category from [Poolside](https://poolside.ai/) and a step forward from their Laguna XS.2 model (released in April 2026). It combines...",
     "hf": null,
     "arena": {
      "elo": 1303.7,
      "votes": 3771,
      "set": "webdev"
     }
    },
    "inclusionai/ling-3.0-tiny:free": {
     "name": "inclusionAI: Ling 3.0 Tiny (free)",
     "context": 262144,
     "created": 1786034890,
     "description": "Ling 3.0 Tiny is a mixture-of-experts model from InclusionAI, with 1.3B active parameters out of 7.9B total. It is designed for responsive agents, instruction following, and multi-turn conversations, with switchable...",
     "hf": null
    },
    "google/gemma-4-31b-it:free": {
     "name": "Google: Gemma 4 31B (free)",
     "context": 262144,
     "created": 1775148486,
     "description": "Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function...",
     "hf": {
      "repo": "google/gemma-4-31B-it",
      "downloads": 10070000,
      "likes": 3493
     },
     "arena": {
      "elo": 1450.9,
      "votes": 5839,
      "set": "text"
     },
     "aa_index": 30
    }
   },
   "takeaway": "First month — nemotron-3-ultra-550b-a5 leads at 9.9. At 9.9, it's worth an on-demand A/B vs GLM-5.2 (multi-model-text-compare) before trialing in any cron.",
   "references": [
    {
     "id": "anthropic/claude-fable-5",
     "name": "Claude Fable 5",
     "role": "top",
     "role_label": "🏆 Top-end",
     "arena_elo": 1507.3,
     "arena_votes": 19390,
     "aa_index": 62,
     "hf": null
    },
    {
     "id": "x-ai/grok-4.5",
     "name": "Grok 4.5",
     "role": "popular",
     "role_label": "🔥 Popular",
     "arena_elo": 1468.4,
     "arena_votes": 14883,
     "aa_index": 56,
     "hf": null
    },
    {
     "id": "zai/glm-5.2",
     "name": "GLM-5.2",
     "role": "workhorse",
     "role_label": "🛠 Workhorse",
     "arena_elo": 1471.0,
     "arena_votes": 24250,
     "aa_index": 53,
     "hf": {
      "repo": "zai-org/GLM-5.2",
      "downloads": 2500302,
      "likes": 4920
     }
    },
    {
     "id": "deepseek/deepseek-v4-flash",
     "name": "DeepSeek V4 Flash",
     "role": "efficient",
     "role_label": "⚡ Efficient",
     "arena_elo": 1435.8,
     "arena_votes": 48561,
     "aa_index": 52,
     "hf": {
      "repo": "deepseek-ai/DeepSeek-V4-Flash-0731",
      "downloads": 954441,
      "likes": 3018
     }
    }
   ]
  },
  {
   "month": "2026-09",
   "label": "September 2026",
   "ts": "2026-09-07T00:06:37+00:00",
   "scores": {
    "minimax/minimax-m3:free": {
     "rank": 1,
     "composite": 10.0,
     "A": 10.0,
     "B": 10.0,
     "C": 10.0,
     "tokens": 1207,
     "tok_s": 57.6,
     "cost": 0.0,
     "dnf": []
    },
    "nvidia/nemotron-3-ultra-550b-a55b:free": {
     "rank": 2,
     "composite": 10.0,
     "A": 10.0,
     "B": 10.0,
     "C": 10.0,
     "tokens": 1896,
     "tok_s": 58.2,
     "cost": 0.0,
     "dnf": []
    },
    "thinkingmachines/inkling-small:free": {
     "rank": 3,
     "composite": 9.8,
     "A": 9.5,
     "B": 10.0,
     "C": 10.0,
     "tokens": 4245,
     "tok_s": 234.4,
     "cost": 0.0,
     "dnf": []
    },
    "thinkingmachines/inkling:free": {
     "rank": 4,
     "composite": 9.5,
     "A": 9.2,
     "B": 9.4,
     "C": 10.0,
     "tokens": 3684,
     "tok_s": 137.0,
     "cost": 0.0,
     "dnf": []
    },
    "dots-studio/dots-3-note-preview:free": {
     "rank": 5,
     "composite": 7.5,
     "A": 8.0,
     "B": 10.0,
     "C": 4.5,
     "tokens": 7250,
     "tok_s": 33.3,
     "cost": 0.0,
     "dnf": []
    },
    "nvidia/nemotron-3.5-lightning:free": {
     "rank": 6,
     "composite": 5.4,
     "A": 0.0,
     "B": 10.0,
     "C": 8.0,
     "tokens": 4125,
     "tok_s": 37.2,
     "cost": 0.0,
     "dnf": []
    }
   },
   "external": {
    "thinkingmachines/inkling-small:free": {
     "name": "Thinking Machines: Inkling Small (free)",
     "context": 1048576,
     "created": 1785443117,
     "description": "Inkling Small is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 12B active parameters out of 276B total. It is positioned as the smaller, more efficient member of...",
     "hf": null,
     "arena": null
    },
    "thinkingmachines/inkling:free": {
     "name": "Thinking Machines: Inkling (free)",
     "context": 1048576,
     "created": 1784325956,
     "description": "Inkling is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 41B active parameters out of 975B total. It is designed for general-purpose reasoning, coding, agentic and tool-use systems,...",
     "hf": null,
     "arena": null
    },
    "minimax/minimax-m3:free": {
     "name": "MiniMax: MiniMax M3 (free)",
     "context": 1048576,
     "created": 1780245374,
     "description": "MiniMax-M3 is a multimodal foundation model from MiniMax. It supports text, image, and video inputs with text output, a 1M-token context window, and is suited for long-horizon agentic work, coding,...",
     "hf": null,
     "arena": null
    },
    "nvidia/nemotron-3.5-lightning:free": {
     "name": "NVIDIA: Nemotron 3.5 Lightning (free)",
     "context": 1000000,
     "created": 1786452751,
     "description": "NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model from NVIDIA, with 3B active parameters out of 30B total. It is suited for high-throughput agentic workloads and specialized tasks that...",
     "hf": null,
     "arena": null
    },
    "nvidia/nemotron-3-ultra-550b-a55b:free": {
     "name": "NVIDIA: Nemotron 3 Ultra (free)",
     "context": 1000000,
     "created": 1780551208,
     "description": "NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...",
     "hf": {
      "repo": "nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16",
      "downloads": 231785,
      "likes": 338
     },
     "arena": {
      "elo": 1426.1,
      "votes": 10700,
      "set": "text"
     },
     "aa_index": 38
    },
    "dots-studio/dots-3-note-preview:free": {
     "name": "Dots Studio: Dots3-Note Preview (free)",
     "context": 512000,
     "created": 1786680361,
     "description": "Dots3-Note Preview is an open-weight mixture-of-experts model from Dots Studio, with 16B active parameters out of 280B total. It is the lightest model in the Dots 3 family and is...",
     "hf": null,
     "arena": null
    }
   },
   "takeaway": "minimax-m3 took the top spot from nemotron-3-ultra-550b-a5. At 10.0, it's worth an on-demand A/B vs GLM-5.2 (multi-model-text-compare) before trialing in any cron.",
   "references": [
    {
     "id": "anthropic/claude-fable-5",
     "name": "Claude Fable 5",
     "role": "top",
     "role_label": "🏆 Top-end",
     "arena_elo": 1507.3,
     "arena_votes": 19390,
     "aa_index": 62,
     "hf": null
    },
    {
     "id": "x-ai/grok-4.5",
     "name": "Grok 4.5",
     "role": "popular",
     "role_label": "🔥 Popular",
     "arena_elo": 1468.4,
     "arena_votes": 14883,
     "aa_index": 56,
     "hf": null
    },
    {
     "id": "zai/glm-5.2",
     "name": "GLM-5.2",
     "role": "workhorse",
     "role_label": "🛠 Workhorse",
     "arena_elo": 1471.0,
     "arena_votes": 24250,
     "aa_index": 53,
     "hf": {
      "repo": "zai-org/GLM-5.2",
      "downloads": 2500302,
      "likes": 4920
     }
    },
    {
     "id": "deepseek/deepseek-v4-flash",
     "name": "DeepSeek V4 Flash",
     "role": "efficient",
     "role_label": "⚡ Efficient",
     "arena_elo": 1435.8,
     "arena_votes": 48561,
     "aa_index": 52,
     "hf": {
      "repo": "deepseek-ai/DeepSeek-V4-Flash-0731",
      "downloads": 954441,
      "likes": 3018
     }
    }
   ]
  }
 ]
}